OxfordHack 2017 project - DeepDoc
The goal of the project was to build an application to help doctors diagnose complicated diseases. Our approach was two-fold: we wanted to use deep learning to process large amounts of quantitative data and classify this as a particular disease, as well as to listen to the worries of the patients themselves.
The project is a medical software that can predict and visualise the likelihood of an individual patient to have each type of cardiac arrhythmia based on their medical data. It also has a Doctor bot for Facebook Messenger that can ask and note some simple symptoms from a patient, as well as their severity and duration.